Ancient Philosophy 13 (1993)©Mathesis Publications

Problems with Vlastos's Platonic Developmentalism

Debra Nails

273

Adherents to developmentalism are eommitted to the interloeking premisesthat Plato's views evolved or developed over his produetive lifetime, and that theehronologieal order of eomposition of the dialogues ean be reeonstrueted withsuffieient eonfidenee to yield a mapping of doetrines to dialogues. With one fur­ther premise, that the earliest dialogues depiet the views of the historiealSoerates, the orthodoxy of Anglo-Ameriean PIatonie studies is off and running.The most influential developmentalist of this generation is Gregory VIastoswhose long ehain of artieles over many years, advoeating and elaborating thesepremises, has been widely aeeepted: 'early' dialogues are marked by Plato'sadherenee to the doetrines of the historieal Soerates, 'transitional' ones by amovement away from 'Soeratie' views and the appearanee of seminal theories,and 'middle' ones by a philosophieal maturity in whieh the eharaeter Soerateshad beeome a mouthpieee for Plato' s very different doetrines. 1

By his own aeeount, VIastos long defended his hypothesis with assumptionand interpretation, albeit an erudite interpretation.2 But Socrates: Ironist andMoral Philosopher ehanges the legaey of Vlastos's proposed ehronology and itsimplieations for Soeratie and Platonie interpretation by systematieally setting outelear and textually grounded arguments for ten theses about the early Soerates,SoeratesE' and ten opposing theses about the middle Soerates (mature Plato),SoeratesM.3 Defying eharges of eireularity, VIastos deelares that he is 'orderingthese dialogues solely by their philosophical content' (his emphasis 1991, 46n2).

1 VIastos 1991, 46n2 credits Ross 1933,7-24 with the basic content of the chronology, and withhaving brought together the arguments of what was already a substantial consensus among his fore­bears by 1924. VIastos refines Ross's early-middle-Iate groups with adjustments within the groups,the most influential of which is the partitioning of a group of transitionals. An anonymous referee forAncient Philosophy rightly points out that the stylometric foundation for Ross's (hence Vlastos's)chronology was laid by Campbell in 1867. The stylometric evidence is complex and controversial; Iprefer to limit myself here to Vlastos's doctrinal claims because they are a well-focused effort todefend the developmentalist position. I have argued elsewhere, however, that previous stylometricefforts (partially excepting Ledger 1989) have been flawed: Nails 1992 and forthcoming. On the moregeneral issue of the defects of developmentalism, see the comprehensive treatment by HolgerThesleff 1982, esp. 7-52, and the wealth of references therein, and, for recent and differently focussedapproaches, Howland 1991, Nehamas 1992, and Kahn 1992.

2 See VIastos 1983a, 513nl0; 1984, 202nl; 1985, Inl; 1988a, 373n39; and 1988b, 89-11l.3 VIastos 1991,46 comments that only a 'schizophrenie' could pursue such different philoso­

phies as he will shortly outline, but Ludwig Wittgenstein and Hilary Putnam provide familiar exam­pIes of philosophers who have held widely divergent views at different times.

274

Fair enough. Even if constructed by an unabashedly circular process-friendscall it 'bootstrapping'-if the structure is sound and has compelling explanatorypower for our understanding of the dialogues, its value should be readilyacknowledged,4 at the very least, the already widespread use of Vlastosian doc­trine designations ('early' and 'middle') and corresponding dialogue groupingsshould gain further justification. But I do not believe the structure is sound, andtherefore I am suspicious of the explanatory potential claimed for VIastos 'ssequence of dialogues. My narrow aim in this paper is to illustrate, althoughmore schematically than the individual issues deserve, that VIastos' s principaldoctrinal distinctions do not hold for the dialogues as he groups them chronolog­ically, that a host of ad hoc arguments, if not special pleading, is required tomaintain his program. If I am right, substantive inferences from some dialogue orother's being thought 'early' or 'middle' should be drawn with cautious restraint,if at alle

The determination of whether the chronological groups are properly describedby the theses involves a straightforward comparison of the sequentiallist of dia­logues to the list of theses. Here then is the familiar set of Platonic dialogues inthe chronological order VIastos proposes (1991,46-47):

GROUPI(elenctic)

ApologyCharmidesCritoEuthyphroGorgiasHippias MinorIonLachesProtagorasRepublic i

(transitional)

EuthydemusHippias MajorLysisMenexenusMeno

GROUP 11

CratylusPhaedoSymposiumRepublic ii-xPhaedrusParmenidesTheaetetus

GROlTP 111

TimaeusCritiasSophistPoliticusPhilebusLaws

Figure 1

The figure's appearance of clarity is unjustified. Vlastos's alphabeticallisting ofthe elenctic and transitional dialogues obscures finer chronological detail that islater deployed in the arguments of his text. He refers to the Gorgias as 'the last ofthe Elenctic Dialogues' (1991, 115n39), and argues that the Euthyphro is laterthan Protagoras, Charmides, and Laches (1990, 14nll). Lysis is earlier than theHippias Major (1990, 14nll) and Euthydemus (1991, 130); Meno is later (1990,9; 1991, 130). Menexenus is n1issing from Vlastos's 1990 (14nll) list of post-

4 In fact, VIastos makes an effort to break the offending circle, and more clearly to distinguishthe doctrines of SocratesE from those of SocratesM' by appealing to the ancient testimony of Aristo­tle and Xenophon, and by articulating a complex conception of Socratic irony to explain persistentdiscrepancies. These are matters for another occasion.

275

elenctic dialogues; apart from noting its terminus post quem of 387, he takes noposition in 1991 on its order with respect to the other dialogues of the transitionalperiod.

The seemingly tidy group II is compromised by a number of exceptions thatfootnotes and later text reveal. SocratesM (i.e., mature Plato) of the Parmenidesis the character Parmenides, not the character Socrates, and SocratesM of theSymposium is the character Diotima, not the elenchus-practicing characterSocrates.5 It is unclear whose views are supposed to be represented by the char­acter Socrates? in Parmenides (except that he 'has been turned for the nonce intoajuvenile inexperienced dialectician', 1991, 85nll) and Symposium. The Sympo­sium presents an additional complication: Alcibiades' s speech is reckoned to be'Socratic' (more than that, it is VIastos 's quintessential source for the Socraticirony supposed to be characteristic of SocratesE' 1991, 33, 47n11). In addition,two of the Phaedo's three biographical scenes, 57a-61c and 115c-118, aredeemed 'Socratic' and therefore available for 'early' plumbing (though the onein between them, 96e-9ge, is deemed autobiographical of Plato and thereforetruly group 11). Finally, VIastos gleans among 'things we would not have learnedfrom Plato if only the Elenctic and Transitional Dialogues had survived' (1991,251): Socrates's facial physiognomy and 'lowly' social origins (Theaetetus 143e,209c, 149a); other supposed biographical data from the same dialogue areignored (e.g., that Socrates was elitist, by thesis VI criteria, sending some aspir­ing students to Prodicus, 151b), but no principle is given for deciding whichstatements merit our belief. With all these qualifications in mind, let us now turnto the theses VIastos proposes.

Although the argument of this paper is directed to only three theses-I, 111, andX-I must quote the entire set because I will explain below the sense in whichthe three I have chosen are more fundamental than the others (1991, 47-49):

IA. SocratesE is exclusively a n10ral philosopher.IB. SocratesM is moral philosopher and metaphysician and epistemologist and

philosopher of science and philosopher of language and philosopher of religionand philosopher of education6 and philosopher of art. The whole encyclopedia ofphilosophical science is his domain.

IIB. SM had a grandiose metaphysical theory of 'separately existing' Forn1Sand of a separable soul which learns by 'recollecting' pieces of its pre-natal fundof knowledge.

IIA. SE has no such theory.

lIlA. SE' seeking knowledge elenctically, keeps avowing that he has none.IIIB. SM seeks demonstrative knowledge and is confident that he finds it.

5 This is implicit in 1991,73-74, where VIastos says that SocratesM 'declares in T22 ... ' whereT22 is entirely Diotima's words.

6 From its earlier formulation (1988b), 'philosopher of education' has been added to whatSocratesE is not, and 'political philosopher' has been removed.

276

IVB. SM has a complex, tripartite model of the soul.IVA. SE knows nothing of this nl0del, which would have unsettled his concep­

tion of moral virtue and undercut his doctrine of the impossibility of incontinence(akrasia).

VB. SM has mastered the mathematical sciences of his time.Vi\.. SE professes no interest in these sciences and gives no evidence of exper-

tise in any of them throughout the Elenctic dialogues.

VIA. SE's conception of philosophy is populist.VIB. SM's is elitist.

VIIB. SM has an elaborate political theory whose ranking order of constitu­tions places democracy with the worst of contemporary forms of government,lower than timocracy and oligarchy, preferable only to lawless tyranny.

VIIA. SE has no such theory. Though harshly critical of political goings-on inAthens, he says that he prefers the city with her laws to any contemporary state.But he leaves the rationale of the preference unexplained.

VIllA & B. Homoerotic attachments figure prominently in the conception oferos in both SE and SN!. But in the latter they have a metaphysical grounding inlove for the transcendent Form of beauty which is wholly lacking in the former.

IXA. For SE piety consists in service to a deity which, though fully supernatu­ral, is rigorously ethical in its own character and in the demands it makes on men.His personal religion is practical, realized in action.

IXB. SM's personal religion centers in communion with divine, but imper­sonal, Forms. It is mystical, realized in contemplation.

XA. In the Elenctic Dialogues SE' s method of philosophical investigation isadversative: he pursues moral truth by refuting theses defended by dissentinginterlocutors. This ceases in the Transitionals: there he argues against theses pro­posed and opposed by hirnself.

XB. In the sequence of dialogues from the Meno through Phaedrus SM is adidactic philosopher, expounding truth to consenting interlocutors. Thereafter themetaphysical theory of the preceding dialogues of the middle period is subjectedto searching criticism by 'Parmenides' and then Socrates, assaying a fresh start,shifts to a new, 'maieutic', mode of investigation in the Theaetetus.

Benign overlap among the theses is obvious. Thesis I is amplified by theses IIand IV (which overlap one another), V, IX, and (to a lesser extent) VII-Vlas­tos' s discussions of which provide textual elaborations of the labels he invokes inthesis I. Indeed, VIastos says he expects the first thesis to unfold through hisaccount of the others (1991, 53); similarly, I hope some of the others will unravelthrough my account of the first. 7 Theses II and III already include the only point

7 But not very far. A systematic and comprehensive refutation of the subordinate theses wouldrequire a contest of texts, the outcome of which would be of interest primarily to those already com-

277

of difference expressed by the halves of thesis VIII; and thesis X offers a moreinclusive account of the conduct of philosophy than does thesis VI. Moreover,some theses (e.g., IX) only occasionally admit opportunities to be tested in thedialogues; and VIastos does not attempt to (re)produce evidence for all his theses(e.g., VII) in 1991, often because he has written extensively elsewhere on theseissues. Nevertheless, he points out that there are far more than ten differencesbetween SocratesE and SocratesM' and pauses to discuss those other distinctionsfrom time to time. Having said that I take theses I, 111, and X to be more funda­mental, by which I mean more inclusive, than the others, I must acknowledgethat VIastos holds thesis 11 to be the most important of the lot. That fact obligesme to return to it after I have made my case.

Thesis I. The Ion violates this thesis, as VIastos admits implicitly when heunblinkingly raises 'the theory of poetic inspiration which he [SocratesE] devel­ops with great gusto in the Ion, alluding to it also in the Apology' (1991, 168).The addition of the Apology is significant because, according to VIastos, 'withoutit we would be left wondering if the theory of poetic inspiration expounded in theIon is pure Platonic invention, without any foundation in authentic Socraticthought' (1991, 288). I find this treatment of the Ion difficult to explain except asan oversight (the more so because the admission that a supposedly elenctic dia­logue examines a non-moral philosophical subject will put the Ion in conflictwith thesis X as weIl).

A more interesting and potent exception to the claim that SocratesE is exclu­sively a moral philosopher-or that Charmides is an early dialogue, to put thedispute the other way around-is the lengthy discussion of whether 'knowledgeof knowledge and non-knowledge' exists (166e-175c). Conceding that the topicis epistemology, in his discussion of the passage VIastos explains that

[i] Socrates's 'arguments sometimes trench on other topics, butthe only theses he investigates elenctically are propositions inthe moral domain'. Besides,[ii] Socrates is compelled to investigate the subject 'onlybecause it was proposed as (an unacceptable) definiens ofsophrosyne; and[iii] he gives up the search when he becomes convinced that itis not likely to get anywhere, confessing that[iv] he has no confidence "in his ability to clear up thesethings" [169a]' (1991, 47-48 n12).

Let us examine the parts of VIastos 's explanation.It is in one sense impossible to refute the claim in [i] because VIastos (1983,

30ff.; reaffirmed 1991, 14) renders it trivially true by defining the method ofelenchus as 'a search for moral truth', effectively precluding its use for other

mitted to the early-middle-Iate strueture as a solution to the Soeratie problem. If my analysis is eor­reet, that very eommitment is ealled into question.

278

maUers. So if Charmides violates thesis I, it violates thesis X perforce, accordingto which SocratesE employs only the elenctic methode Vlastos's [ii] combineswith [i] in seeking to undermine the epistemological discussion by supposing thatSocrates was merely following where his interlocutor led. In other words, wereaders ought to take SocratesE' s foray into epistemology as an action in the ser­vice of moral philosophy. This interpretation, applied seemingly harmlessly inthe Charmides, is a weasel that would nose its way back in at will, wreakinghavoc on more of the Vlastosian structure. For example, all the ontology of theRepublic has too often been seen as 'in the service' of Plato' s political philoso­phy. Arguably, the entire point of the Republic is a moral one: to demonstratethat the slave who found Gyges' ring was, ultimately, unhappy and worse off.And perhaps the moral philosophical discussions of the Gorgias are ancillary tothe question, What is rhetoric? There is no essential difference between thesepositions and the claim that the epistemology of the Charmides is subordinate toits ethics. 8

VIastos 's point [iii] is false: the Socratic confession he cites occurs only aquarter of the way through the epistemological discussion~ thus, far from givingup the search, as VIastos claims, Socrates is represented by Plato as digging in.And Socrates's comment [iv] about his inability 'to clear up these things' paral­Ieis perfectly the remarks VIastos himself will be quoting in defense of his thesis111 to demonstrate that SocratesE disavows knowledge (see below). If we aresoon to be exhorted to interpret such remarks not literally but as forms of com­plex irony, why not now?

Even the Apology raises epistemological problems. VIastos points out, andcriticizes others for failing to recognize, that the Apology (20-21) is the dialoguethat clarifies what kind of knowledge Socrates means to be disavowing when heclaims, as he so often does, that he has no knowledge (1988b, 99-100~ 1991,238­240): while 'more than human knowledge' is the prerogative of the god, knowl­edge or wisdom of the 'human kind' Socrates is willing to admit he hase VIastosadds, 'Precisely how he wants us to understand the "human knowledge" heavows and the "more than human knowledge" he disavows he does not explain:he could hardly have done so without plunging over his depth in epistemology'(1988b, 100n69). We are free to suppose instead that Socrates does not explainbecause he could hardly have done so without plunging his jurors into an expla­nation over thelr depth. VIastos 's blinkered Socrates is difficult to reconcile witha person who, especially according to VIastos, makes such epistemological dis-

8 Whatever the differences between SocratesE and SocratesM on moral philosophy, one thingthat is the same throughout is the continuity of ethics and epistemology. In the context of the discus­sion of Protagoras, VIastos will say that SocratesE is not an ontologist because he does not reflect onthe ontology he holds; presumably we are to infer that SocratesE does not reflect on his epistemologyeither (since it is claimed he is no epistemologist). Contemporary philosophers seem quite unable todiscuss SocratesE' s moral philosophy without bringing in its epistemological underpinnings, so thereis ample reason to doubt that Socrates himself skated merrily along on the surface of issues concern­ing how we ought to live.

279

tinctions as the Apology reports.SocratesE also moves beyond moral philosophy in the Protagoras, with its

ontological dispute over whether 'man is the measure of all things ... '; not thatProtagoras says it in so many words, but his view that virtue is necessarily con­ventional rests on that ontologieal premise. More centrally, there is the discus­sion of being and becoming beginning at 339, and the discussion of appearanceand reality at 356. Throughout, Socrates' s opposition to Protagoras' s point ofview indicates that Socrates has a competing ontologieal position of his own.This VIastos concedes. The crux is VIastos 's assertion that Socrates, who 'hasthis ontology' (his emphasis), is 'no ontologist' (1991, 58).9 SocratesE' accord­ing to VIastos, does not reflect on his own ontology, takes the reality of the formsfor granted as a New Yorker takes for granted trees and stones. This claimdeserves a closer look. Socrates and the New Yorker both take trees and stonesfor granted, which is not surprising, given how the human senses function-butforms? The problem arises in Socrates's having an ontology of forms whichVIastos cannot have both ways: either believing in forms was common (as theParmenides has it), or it was uncommon (resulting in Socrates's having so manyinterlocutors who fail to give hirn the universal definitions he seeks). If it wascommon, Plato should receive no fanfare for originating the theory; if uncom­mon, then either Socrates developed them on his own (so he is quite an ontolo­gist, even if his forms are not separately existing), or he learned them fromsomeone else, in which case it is very difficult to believe he did so withoutinquiring into their nature. VIastos could respond that the forms were just apartof the furniture of Socrates' s mind, like trees and stones and his daimonion per­haps; and that Plato's development of a theory of "'separately existing" Forms' iswhat deserves credit.

But VIastos takes another tack as wen, explaining that SocratesE does notpress Protagoras on the subject of moral relativism because to do so would leadbeyond the moral topics of the moment and 'Socrates is to be kept down to therole of single-minded moralist' (1991, 62n68). 'Kept down?' Protagoras is notransitional dialogue, so Plato and Socrates are supposed to be of one mind at thisstage. Who must keep Socrates down, and why? There is a more important issueat stake than the mere interpretation of VIastos 's peculiar remark, though this isnot the place to develop it properly: It can be said of an the aporetic dialoguesthat they need a theoretical structure to ground them, that they need the episte­mology and ontology VIastos ascribes only to SocratesM. Plato writes such ajus­tification in, for example, the Republic. But it was not necessary to tack on thoseother nine books, so to speak, in every case; the dialogues considered individu­ally have philosophical and aesthetic unities that Plato appears to have valued,

9 In a note to the passage in which VIastos discusses the Protagoras, he acknowledges his ownformer denial (thirty years back) of the plausibility of his current position, and credits Woodruff 1982with making a better case against it than he (VIastos) was able to make at the time. Thesleff hasbroadened my appreciation of the extent of ontological discussion in the dialogue, for which I amgrateful.

280

and appropriately, beyond a completeness that would necessarily have broughtrepetition with it.

If the subject of the Gorgias is rhetoric (Benardete 1991), if the subject of theEuthyphro is religion, then thesis I is violated in those instances as weIl, but Iwould not advocate either case. Nor would I deny, however, that interesting anduseful philosophy can begin with the assumption that the subject of these twodialogues is other than moral philosophy.

Thesis 111. With this thesis that SocratesE disavows knowledge, VIastos(1990; 1991, 82-86) seeks to establish that SocratesE never committed the so­called Socratic fallacy that appears for the first time, he says, only in the transi­tional dialogues. Robinson (1953, 50), according to VIastos, incorrectly inferredits commission in the Protagoras (360e-361 a) where Socrates, having just beenaccused of merely trying to win an argument and forcing Protagoras' s replies,defends himself, 'I have no other reason for asking these things than my desire toanswer these questions about virtue, especiaIly what virtue is in itself. For I knowthat if we could get clear on that, then we would be able to settle the questionabout which we both have had much to say, I-that virtue cannot be taught,you-that it can' (Lombardo and Bell tran.). Robinson's argument is containedin a more general 'What is X' discussion that cites Protagoras 360e parentheti­cally without further comment, so VIastos (1990, 10-11) must first build a casefor Robinson, then refute it. The argument he constructs yields that knowingwhat X is is a necessary condition for knowledge about X in the Meno, but only asufficient condition in the Protagoras (and, by extension, in the other early dia­logues)-where the latter is inadequate for acharge of Socratic fallacy. If theargument did in fact turn on the interpretation of two snippets of text, with theoutcome requiring only that Plato be cognizant of, and deliberate about repre­senting, the difference between the two formulations, then I might be amenableto VIastos 's interpretation. But Robinson' s position is very nluch broader thanVIastos lets on. Robinson (1953, 51) builds his own argument not only on thetransitional dialogues, Meno (71, 86d-e) and Lysis (223b), of which VIastosapproves, but on the elenctic ones, Euthyphro (6e), Republic i (354c), and Laches(18ge-190a),10 concluding, 'In fact, the impression vaguely given by the earlydialogues as a whole is that Socrates thinks that there is no truth whatever aboutX that can be known before we know what Xis.' Try as I might, I cannot get allthe passages to conform themselves to VIastos' s sufficient-condition formula­tion. 11 Plato, at least, seems to have been unaware of it.

The passage at Republic i (354c) is worth pausing over, for it is not an instancethat VIastos hopes he has explained adequately with his account of necessary andsufficient conditions. No, the maintenance of the thesis that SocratesE never

10Irwin 1977, 40 cites 190c to the same effect, though Vlastos 1990, 14n8 rejects that too.II Others have seen the problem in Charmides although Santas 1972, 138 disproved its occur­

rence there, according to Vlastos.

281

commits the Socratic fallacy requires in this case lopping off the tail of Republici, after 354a11 (1985, 26n65). This he explains as follows, 'That this closingparagraph could not have belonged to the original dialogue is shown by the con­tradiction between Socrates' saying at 354cl-2 that, since he does not know whatjustice is, he cannot know whether or not it is a virtue, and his earlier claim tohave shown that no one could fail to know (hence Socrates could not) that justiceis a virtue (351a3-6)' (1990, 15n31). But why take so radical a solution? WhenVIastos is at pains to show that Socrates is not a crook (in his chapter 'DoesSocrates Cheat?'), he is ingenious with his suggestions. To keep Socrates fromcheating Polus, and simultaneously keep Plato from 'a crisis in his own charac­terization of his teacher', VIastos says helpfully, 'Surely it is simpler to supposethat he [SocratesE] is hirnself unaware of the fallacy' (here, an inferential fallacyin the Gorgias). A little later, VIastos is willing to excuse Plato for being'messy': 'Plato is simulating a live conversation. When people are arguing ontheir feet not all of their arguments can be expected to come through in apple-pieorder' (1991, 148 and n68). The problem for someone who would read Platothrough VIastos 's lens is when to apply which principle: do I ditch the passage,plead Socratic ignorance, praise Platonic realism-or retreat to the last trench,irony?

Gf Gorgias, VIastos grants that SocratesE seeks knowledge, not just true belief(1991, 14).12 The question is whether he ever finds it and says he has found it­which he must not if VIastos 's thesis 111 holds for that dialogue. SocratesE mustkeep avowing that he has no knowledge. Yet (in VIastos 's own translation)SocratesE does claim knowledge (1991, 4-5 and n16): 'When told in the Gorgias(473b) that it would not be difficult to refute his thesis, Socrates retorts: "not dif­ficult, Polus, but impossible; for what is true is never refuted" (T 0 yap aA:fJeE~

OVOE7TOTE EAEyXETaL). Later (47ge) he asks: "Ras it not been proved [a7TOOEOELK­TaL] that what I said was true?'" 13 Re even refers to his position as 'clampeddown and bound by arguments of iron and adamant', though he immediatelyturns back, 'my position is always the same: I do not know how these maUersstand' (508e-509a).14 VIastos could of course point out that Gorgias is the last ofthe elenctic dialogues before the transitionals, and that the certainty characteristicof Plato' s maturity is beginning to creep in. And that is just what he does in1983c (74) wrestling virtue from contradiction: the assumption that VIastos saysSocratesE defends (see discussion of thesis X below) is characterized as both anoutgrowth of the conversation and 'Plato' s present to his teacher', heraldingPlato' s loss of faith in the elenctic method.

12 Nehamas 1992, 165-166 criticizes VIastos for relying too much on the Gorgias generally.13 The passage quoted sounds very like what the character Socrates says to Agathon at Sympo­

sium 201c-d: 'No, it is Truth, my lovable Agathon, whom you cannot contradict: Socrates you easilymay' (tran. Lamb: Ov }J.EV ovv rfl aA:ry8ELa ... ovva(J"aL aVTLAEyELV, E7TEL ~WKpaTEL yE OVOEVxaAE7Tov); but VIastos allows no comparison here: the character Socrates is of indeterminateearly/middle status in the Symposium because Diotima has been designated SM'

14 VIastos 1990,4 offers two more such instances in the Gorgias: 472c-d and 505e.

282

By 1991, however, VIastos had developed a more comprehensive approach tothe subject of his third thesis, explaining Socrates's about-face in the Gorgiaswith recourse to the notion of paradox. I must back up slightly: VIastos (1991,82-86) identifies three features of the disavowal of knowledge. The first two­that Plato treats it as a notorious trait, and that Plato never n1akes Socrates' s inter­locutors ask hirn for an explanation of it-need not concern uso But the third,introduced as 'the most paradoxical aspect of the disavowal, its unique, abso­lutely unparalleled, feature', is that 'it may be voiced at the conclusion of anentirely successful elenctic argument in which Socrates has to all appearance,proved his thesis to the hill', whereupon Gorgias 508e-509a is cited as an exam­pIe. Moreover, we are told this never happens after the transitional Meno. ButVIastos notes that two middle Platonic passages might seem to conform to thethird early feature, so he discusses each. The first (RepubUc 445b) is not anexample of the required type, but the second (Theaetetus) is cfUcial to the claimof thesis III that 'SocratesM seeks demonstrative knowledge and is confident thathe finds it.'

In the Theaetetus, though dead last of the group II dialogues, SocratesMsounds like SocratesE in disavowing knowledge, 'I am sterile in point of wisdom.. .1 have no wisdom in me.. .1 am, then, not at all a wise person myself, nor haveI any wise invention ... ' (150c-d, Fowler tran.). But none of these pronounce­n1ents contradicts the letter of VIastos 's assertion because they occur early on,not 'at the very moment at which he has produced evidence which appears tobeUe it [the disavowal]' (his emphasis).15 VIastos concedes that Socrates'emphasizes he has not found the answer to the "What is the F?" question aboutknowledge' at the end of the dialogue, but insists that Socrates is not disavowingknowledge so globally as he did in the beginning. VIastos quotes Socrates, 'Ihave none of the knowledge possessed by all the great and wonderful men of thepasl' , and adds: 'he says nothing here or hereafter which appears to cast doubt onwhat he has come to know in the course of it [the inquiry]'. Narrowly, 'here orhereafter' is correct-but the clause immediately before the one quoted doesindeed cast doubt: Socrates had just asked, 'Then does our art of midwiferydeclare to us that all the offspring that have been born are mere wind-eggs andnot worth rearing?' and Theaetetus had answered that it does. Socrates had con­tinued, predicting on the basis of their conversation, 'you will have the wisdomnot to think you know that which you do not know. So much and no more my artcan accomplish; nor do I know aught of the things that are known by others, thegreat and wonderful men who are to-day and have been in the past' (Fowlertran.). What VIastos does not quote echoes what he calls 'the clearest statementin Plato' of the disavowal of knowledge (1991, 84): 'I have no knowledge, nei-

15 Besides, Vlastos's chronology has the Parmenides immediately preceding the Theaetetus; heattributes the dearth of positive doctrines in the latter to Plato's 'declining to meet head-on' the'formidable difficulties' raised in the Parmenides (1991, 85). VIastos also points out, and correctlyenough, that Socrates does not repeat his disavowal of knowledge at l87a, when the first of threecon1petitor definitions of knowledge has been discarded.

283

ther do I think I have any' (Apology 21d, VIastos tran.). Here, and again below,the Socrates of the Theaetetus will give us reason to collapse the early/middledistinction.

There are other examples of group 11 dialogues in which SocratesM disavowsknowledge, even 'at the very moment' when he has produced seemingly conclu­sive evidence: the clearest example appears in an elenctic section of the Cratylusthat begins and ends with just such disavowals: 'I do not know what the truth isabout such maUers' (384c) ... 'You forgot what I said a while ago, that I did notknow, but would join you in looking for the truth' (391 a). Even in the Phaedo(91 b), where VIastos finds SocratesM so full of positive doctrines, Socratesclaims no certainty for his arguments and, referring to his impending death,remarks 'this ignorance of mine will not last'; in the Phaedrus (235c) he says, 'Iam conscious of my own ignorance' (all Fowler tran.).

Thesis X. According to VIastos 's account of Socrates' s method, SocratesEpractices an adversative, elenctic method that is supplanted in the transitionaldialogues by a somewhat similar method except that Socrates hirnself proposesthe theses he then refutes; in stark contrast, SocratesM is at first didactic, thenmaieutic. I will show that these divisions are not so neat as they may at firstappear, neither for the elenctic dialogues, nor for the didactic and maieutic ones.There is a simple point to dispense with first. SocratesE is not obliged always touse the elenctic method. Right off the bat, neither the Apology nor the Crito iselenctic, though there is some question-and-answer in both. In the latter, and inthe Gorgias, SocratesE sounds didactic, like SocratesM' delivering protrepticspeeches (quite a few between Gorgias 464b and 527e), which VIastos readilyacknowledges, excusing them because they occur 'only after hard-won elencticargument had established the great truths which the interlocutor is then exhortedto take to heart'. When SocratesE makes protreptic speeches in the Protagoras(310a-314b), he is absolved by VIastos because he has provided a 'curtain raiserto the elenctic drama it precedes' (1991, 116-117 n46).16 So protreptic is judgedan acceptable SocraticE practice if it follows or precedes elenchus-just not if itreplaces it.

The discussion of elenchus to which VIastos has always referred his readers(1983b, 30ff; cf. 1991, 14, 113-114) begins: 'Socratic elenchus is

[i] a search for moral truth[ii] by adversary argument[iii] in which a thesis is debated only if asserted as theanswerer's own belief,[iv] who is regarded as refuted if and only if the negation of histhesis is deduced from his own beliefs.'

But there will prove to be serious difficulties with all but the fourth point.Thesis I has already described SocratesE as exclusively a moral philosopher;

16 Similarly, the development of theory in the Ion preceded the elenchus.

284

now thesis X [i] limits the use of the elenctic method to the search for moraltruth, forming a constricting circle. If the definition of 'elenchus' is permitted toremain so narrow, the dialogues of group II will prove trivially non-elenctic,even if segments of dialogue are otherwise demonstrated to fit the description ofelenchus, as has already been found with Charmides and Ion. VIastos elaboratesof the Ion, 'In explaining the poet to us, Plato lets Socrates speak like a poet forthe nonce. But he does not make Socrates abandon his customary elenctic role onthat account' (1991, 287). Since he had already said, 'Socrates uses the elenchusexclusively in the pursuit of moral truth, remaining fronl first to last a single­minded moralist' (1991, 14), we must jettison one statement or the other. For tworeasons, I suggest the minimal amendment of changing 'exclusively' to 'almostexclusively' for the nonce: (a) we get no philosophical mileage from the stiffercriterion, given thesis I; (b) there is a large literature already distinguishing dif­ferent varieties of Socratic method (cf. n18 below), but these are issues of whatprocedure is practiced. VIastos 's restriction of elenchus to moral philosophy israther an issue of the domain in which a procedure is practiced, but no argumentis given to support such a change of focus.

If the elenctic method is thus freed of constraint, Charmides and Ion, treatingepistemology (in part) and philosophy of art, respectively, cease to violate thesisX. But group II dialogues with elenctic sections do violate it and disprove Vlas­tos' s assertion that the 'hybrid' Meno marks the final appearance of the elencticmethod 'because Plato hirnself has now lost faith in that method' (1991, 117). Inthe Cratylus, SocratesM uses the elenchus to disabuse Hermogenes of a viewabout the philosophy of language that the interlocutor sincerely believes, pro­poses and defends: 'I think no name belongs to any pal1icular thing by nature, butonly by the habit and custom of those who employ it and who establish theusage' , he says (384d). By the end of the standard elenctic discussion, Socratesconcludes, 'Then, Hermogenes, the giving of names can hardly be, as you imag­ine, a trifling matter, or a task for tritling or casual persons: and Cratylus is rightin saying that nanles belong to things by nature and that not every one is an arti­san of names' (390e, Fowler tran.). There is one example of the elenctic methodin the Phaedo, though it is admittedly truncated and reversed; at 92a-e Socratesdeduces the negation of Sirnrnias' s belief that the soul is like the harrnony of alyre from Simmias's own belief that knowledge is recollection, through whichSimmias had been led dialectically earlier in the dialogue (though, even then, itwas not new to hirn). Socrates? of the Symposium leads Agathon through anelenctic refutation of the claim that Love is beautiful (199c-201d). Parrnenides(SocratesM) leads Socrates? through several elenctic rounds in the first section ofthe dialogue (Parmenides 130e-134e), and the elenchus returns yet again in thePhilebus of group 11I.17 (The nlethod of the Theaetetus is special enough that it

17 This was argued by Davidson 1985, but VIastos 1988b, 109n115 acknowledged that the anal­ysis was correcL Although Davidson was not the first to point this out (see e.g. DJ. Allan' s Introduc­tion to Stenzel 1940, viii) but the case needed badly to be made again after many years in oblivion.

285

will pay us to postpone considering it untillater.)Vlastos's second requirement of the elenchus is that [ii] argument be adversa­

tive, that interlocutors themselves provide the theses to be examined (1983b, 57­58) which they do not in Lysis, Hippias Major, and Euthydemus, where VIastosfinds the interlocutors too naIve, too stupid, and too clownish, respectively, forthe method to work. VIastos 's SocratesE is a one-trick pony, an elenchus waitingto happen 'day in day out' (1991,115) but Plato's Socrates (and the historicalone) may weIl have been more versatile. 18 VIastos rules out situations in whichSocrates hirnself introduces subjects because he suspects them to be the beliefs ofhis hearers, for example when an interlocutor has studied with some sophist orother and is reluctant, even ashamed, to defend a sophistical thesis againstSocrates' s attack. The youngsters in the Lysis, for example, are quick to say thattheir teacher is Miccus, whose reputation as a sophist Socrates knows. CouldPlato have desired to illustrate what a philosopher, or what SocratesE' might dowhen confronted with interlocutors unable-for whatever reason-to engage inthe elenchus effectively? I cannot see why not. VIastos can. He says SocratesEwas keenly aware of his mission 'to live philosophizing, examining [hirnself] andothers' (Apology 28e, VIastos tran., his emphasis), and that the method of thetransitionals would have enabled hirn only to examine hirnself, a mere partial ful­fillment of his mission; therefore, VIastos concludes, the transitionals with their'half Socratic' SocratesE must signal Plato' s having realized the elenctic methodwas feckless, a realization SocratesE never experienced. But it is often the case inlife that others allow themselves to be examined only indirectly, or are incapableof elenctic argument; the presence of this attitude or incapacity in potential inter­locutors, presumably, would not have absolved Socrates of the responsibility toexamine them, so I cannot draw the same conclusion VIastos does, even if I granthis premise.

There is a more narrow Vlastosian approach to the issue of the transitionalsand the fact that they fail to instantiate the condition of [ii]. He says that a prob­lem came to light after the Gorgias that persuaded Plato that the elenchus was'highly fallible' (1991, 115) and rested on 'frightfully strong methodologicalassumptions' (1983c, 74): namely, 'Anyone who ever has a false moral beliefwill always have at the same time true beliefs entailing the negation of that falsebelief' (1983b, 52). This realization caused Plato to ditch the elenchus; and whatPlato could not trust, he would not put into Socrates' s mouth. Consequently, thenext three dialogues to appear after the Gorgias (Lysis, Hippias Major, and

18 The position I am treating so glancingly here, that Socrates suited his methods to the psyches

of his interlocutors, is brilliantly developed and defended by Teloh 1986. His nlethodological analy­sis, even more subtle and rich than Robinson' s, explores the appropriate variety of SocratesE' s meth­ods. Teloh discusses the 'benign elenchus (refutation) and psychagogia [used] on Lysis ... becauseLysis is not stuffed with either purported knowledge or false beliefs' (74), in contrast to the eristicused on Menexenus to force him to the point of aporia; the boys' psyches required and received dis­parate approaches from Socrates. Examining other 'early' dialogues, Teloh reveals similar differ­ences in Socratic technique, so the diversity of approach is not limited to 'transitional' dialogues.

286

Euthydemus) were conspicuously lacking in elenchus. But this is a puzzling posi­tion because the three non-adversative dialogues do nothing either to challengethe faulty assumption or in any other way to address the specific problem thatVIastos says had surfaced in the Gorgias. If in the transitional dialogues Platohad chosen interlocutors of equal ability to those of the early dialogues and yetshown SocratesE neglecting to perform the elenchus, the reader might havethought to wonder why; but in choosing such plausibly non-elenctically inclinedcharacters, Plato strengthens the view that different personalities simply requiredifferent types of examination. The ground falls away for VIastos 's inference thatthe absence of adversative argument implies the transitionals were composedlater than the Gorgias.

VIastos makes Theaetetus the final group II dialogue, but that requires a coupleof methodological contortions worth considering independently. Robinson(1953, 204) refers to the Theaetetus as 'so elenctic that it has been mistaken foran early dialogue', but VIastos 's redefinition of the elenchus puts the dialogue ina class by itself. He says that Theaetetus offers theses to be examined, but thatSocrates only refutes them after having first elaborated them himself: 'what herefutes in the course of the long-winded argument that follows is his own imagi­native construct, with which he chooses to saddle his docile interlocutor' (1991,266). But the shell of the elenchus is still intact: SocratesM refutes a thesis byeliciting additional premises that prove inconsistent with the original thesis. Thatmakes the Theaetetus more like the elenctic dialogues than like the transitionals,and much more like the elenctics than like the didactic dialogues of the rest ofPlato's mature period; yet VIastos shuns the term 'elenchus' altogether, callingthe 'new' method maieutic. The content of the Theaetetus, so suggestive of theo­ries to be introduced in the Sophist, may understandably make it tempting to finda rationale for distancing its form from that of the early dialogues, but Socrates' smethod is a very poor ground indeed for giving Theaetetus late group II status.

VIastos reinforces his view that Theaetetus is correctly positioned chronologi­cally by agreeing with Burnyeat (1977) that the image of Socrates as a midwife isa Platonic invention (1991, 16). It is true that the images of labor and birth areconfined to group II dialogues,19 but Sider (1991) has introduced persuasive evi­dence that lines 633 sqq. of Aristophanes' Clouds (originally produced in 423)was staged as abirthing scene, evidence so conclusive that it is hardly possible tomaintain any longer that Platonic invention accounts for the image of Socrates asa midwife.

Another problenlatic aspect of Vlastos's account of elenchus is [iii] that itdemands that the interlocutor not only provide the thesis to be discussed, butbelieve in the proposed thesis as weIl. VIastos (1991,14, 123n69; 1983b, 35-36)

19 VIastos 1991,78-79 quotes, 'The true lover ofknowledge ... may give birth to intelligence andtruth, may know and truly live and be nourished and thus find release from labour-pains' (Republic490a-b); and [the philosopher] 'will give birth to true virtue' (Symposium 212a). Diotima's(SocratesM's) speech is full of allusions to begetting and birthing (cf. 206c and 209).

287

makes clear why he wants Plato to have used a sincerity criterion ('Say only whatyou believe' 1991, 111 n21) in the elenctic dialogues: 2o 'its operative force was toexclude debate on unasserted premises' (his emphasis)-which would be animprovement on the previous brand of dialectic known to the history of philoso­phy and attributed to Zeno. 1 believe SocratesE gave only lip-service to such acriterion, but 1 do not mean that as an insult, as 1 will clarify after examining thefour particular cases in which VIastos teIls us SocratesE observed this principle.

Having noticed that SocratesE seems often 'complicated, devious, cunning'(1991, 133) toward his interlocutors, VIastos seeks a means of determining whenthe reader can be certain Socrates remains 'free of resort to deceit'. The answer isimmediately forthcoming, 'when arguing seriously' (his emphasis 1991, 134). Ifthat seems to verge on non-falsifiable, hold on; VIastos offers an example of con­versation that he calls 'the touchstone of Socratic seriousness': a passage fromthe Gorgias in which Socrates admonishes Callicles, 'you mustn't think that youmay play with me and say whatever comes into your head, contrary to your realopinion, nor, conversely, must you think of me as jesting. For you see what ourdiscussions are all about-and is there anything about which a man of even smallintelligence would be more serious than this: what is the way we ought to live?'(500b-c, VIastos tran. and emphasis). According to this touchstone example,when Socrates raises the issue of how we should conduct our lives, we can becertain he will not be deceitful. What is terribly wrong with the example is thatSocrates does not mean it, is not serious at all: Callicles is refusing within onepage (501c) to cooperate, saying he will agree with what Socrates says just to getthe argument over with. Socrates does not respond by insisting his interlocutorsay what he believes, but by carrying on regardless, and at length. When Callicleslater asks ruefully, 'And you are the man who could not speak unless somebodyanswered you?' Socrates can only reply, 'Apparently 1 can' (519d-e, Lamb tran.)This makes a mockery of thesis X: the touchstone of seriousness itself is anexample of deceit in its context.

Nor does Socrates mean it when he teIls Protagoras, 'I won't have this. For itisn't this "if you wish" and "if you think so" that 1 want to be refuted, but you andme. 1 say "you and me" for 1 think that the thesis is best refuted if you take the"if' out of it' (Protagoras 331c, VIastos tran.). VIastos does not deny that Pro­tagoras is allowed straightaway to abandon the sincerity requirement (233b-c).Instead, he provides a reason for the indulgence granted the sophist, namely,Socrates allows hirn to 'save face' when losing the argun1ent (1983b, 37-38). 1count face-saving among the least respectable excuses for dodging the truth in aserious discussion of the way we ought to live, so 1count the exchange with Pro­tagoras as another strike against the importance, let alone necessity, of a sinceritycriterion.

20 He attributes the view to Robinson (1953, 15-17) as weIl. Although this is not the plaee todevelop Vlastos's wider eoneerns, Vlastos's preoeeupation with other issues of inteIleetual honesty­Soeratie irony, and Soeratie and/or PIatonie deeeit (1991,21-44, 132-156)-that he treats extensivelyand independently, deserve attention in connection with Soerates's method.

288

VIastos (1991, 249) says that, in the first book of the Republic, Socrates'repeatedly insists on the observance of the "say what you believe" rule', citing346a, 349a, and 350e. The text does not bear out his claim. After Socrates hasshaken his finger at Thrasymachus, 'My good man, don't answer contrary toyour real opinion (7Tapa oo~av), so we may get somewhere' (Republic i, 346a,VIastos tran.), he takes it back only three pages later: "'For now, Thrasymachus,I absolutely believe that you are not 'mocking' us but telling us your real opin­ions about the truth." "What difference does it make to you," he said, "whether Ibelieve it or not? Why don't you test the argument?" "No difference," said I, "buthere is something I want you to tell nle in addition'" (349a-b, Shorey tran.).Besides, Socrates has soon browbeaten the sophist into submission. WhenThrasymachus warns hirn that he intends to answer with approval and nods, asone would 'for old wives telling their tales', Socrates retorts 'No, no ... notcounter to your own belief'. But when Thrasymachl1s shoots back, 'Yes, toplease you ... since you don't allow me freedom of speech' (350e, Shorey tran.),Socrates feigns no more reluctance, and Thrasymachus plays the perfect yes-manto the end. VIastos (1991, 249) says the yes-men enter only at book 2, with adver­sative argument characterizing all of book 1, but, in fact, it is SocratesE who putsa stop to Thrasymachus' s engaged participation.

While I do not doubt that, in the Crito (49c-d), when Socrates teIls his friendnot to agree to anything -contrary to his real opinion (7Tapa oo~av), he means itsincerely, I have no evidence whatever that Socrates holds others to the samehigh standard. On the contrary, SocratesE merely pretends to demand seriousnessin the instances outlined above, and the action of the dialogue exposes his lack ofconviction in each case. Nor is it ignoble that this should be so: not only is therevirtue to Thrasymachus' s suggestion that Socrates examine the argument, not theman; not only is the dialectical exhibition useful to others in the crowd (and toPlato' s readers),21 regardless of its lack of effect on the interlocutor; but one ofSocrates's finest attributes is his ability to adjust his conversation to various indi­viduals.22 And these three claims are not mutually incompatible, admitting asthey do something uncontroversial in life, that the conduct of philosophy admitsopportunities at a great number of levels because situations and people differ sowidely. VIastos and I see Socratic versatility in the conduct of philosophy com­pletely differently. Considering Socrates's behavior toward Callicles, Protagoras,and Thrasymachus, I cannot believe interlocutor seriousness is a touchstone ofSocratesE's method; consequently, I do not believe that debate on unassertedpremises is excluded in the dialogues; so I cannot conclude with VIastos that thiswas the historical Socrates' s contribution to the history of dialectical argument(1991, 14).23

21 McPherran 1986 moves convincingly beyond the points I mention here.22 Teloh 1986 argues this point carefully for the dialogues he calls early (which include dia­

logues VIastos calls 'transitional').23 For another view of what makes Socrates' s method an advance beyond that of Zeno, see Irwin

1992, 68-69.

289

Thesis 11. I promised to return to this thesis, that SocratesM holds a theory ofseparately existing forms and a doctrine of recollection alien to SocratesE,24because VIastos considers it 'the most powerful of the ten' (1991, 53). In the fol­lowing sense it iso From the time of Aristotle, but largely because of Aristotle' sinsistent focus on the topic, scholars have viewed the dialogues' differentaccounts of the forms as mirroring 'development' in Plato' s theory; from thenineteenth century, it has been popular to map that perceived development (andsometimes development on other issues) onto the dialogues, producingsequences believed to reflect the order in which Plato composed them.25 VIastos,late to the fray, could build on the progress of his forebears, especially Ross, byrefining the account of the dialogues of the early period with divisions into elenc­tic and transitional. Since his chief principle for ordering the dialogues remainsthe extent to which the theory of forn1s appears in each, it is not surprising thatthesis II results in a relatively good fit. 26 The mapping procedure is a circularone, as extensively noted in the literature, but that would not be damning if thecircle were very wide, or if some of its elements were grounded outside the cir­cle. But this is exactly where we began: a proposed sequence of dialogues was tobe defended by a bold extension beyond the traditional reliance on the theory offorms into other key areas of Platonic philosophy. And that is the project Ibelieve has failed. Even if I accept thesis II in its entirety, since I cannot acceptVlastos's other theses and thereby gain a comprehensive interpretive tool, I amstill in no position to infer anything whatever about the order of composition ofthe dialogues, nor even about what doctrines, if any, the historical Socrates and

24 It seems to me, but I will not argue here, that anamnesis is not nearly so forceful a presence inthe dialogues as the theory of forms and that putting them together in one thesis is unfortunate. Thereare only two passages about anamnesis long enough to be considered accounts of the theory, Meno(85c-86b) and Phaedo (72e-76e), and both are presented with doubts intact: Socrates admits uncer­tainty (KaL Ta J1EV yE aAAa OVK av 7Tavv V7TEp TOV AOyOV enLaXVPL(]"aiJ1YlV, Meno 86b), and empha­sizes that the immortality of the soul (a key premise) is unproven (Phaedo 63c and 114d).

25 Thesleff 1982, 40-42 and n9 discusses the history of different positions taken on that oneissue, providing about three dozen citations from a variety of perspectives. Of those who claim thatone can infer the theory of forms from the Socratic dialogues (those of Vlastos's group I), he listsShorey 1903, Stenzel 1916 and 1917, Regenbogen 1950, Crombie 1962-1963, Allen 1970, Martin1973, Findlay 1974, and Gundert 1977. And of these, VIastos 1991, 59n52 mentions the attractive­ness of Allen's version (without of course approving).

26 I will not here take issue with his proposed sequence (although many others have entered thatlegitimate arena for debate). Apart from the variety mentioned in the previous note, VIastos points outthat Dodds 1959, 375 inferred foreshadowings of the theory of forms in the myth that concludes theGorgias, but is unconvinced that Dodds' analysis is sound. Kahn's (1981, 318; cf. 1988) developmentof the notion of prolepsis in the dialogues leads hirn to argue for 'the anticipation of the Form ofBeauty or Form of Good in the Lysis'. But anticipation is something short of what would be requiredto discredit Vlastos's thesis 11. As Thesleff (1982,40-42 and nn. 2-16) has demonstrated in admirabledetail, however, when scholars have tracked other issues through the dialogues, other chronologicalsequences have resulted. Examples among the fifteen such issues he cites (providing bibliography)are: the virtues, education, logic, and method; some of the issues invite comparison to the topics ofVIastos 's other theses. The selected 132 proposed dialogue sequences Thesleff (1982, 8-17) recountsbetween Tennemann in 1792 and Kahn in 1981 is sobering.

290

Plato may have held. To make either of those leaps, I would need to accept Vlas­tos' s description of Plato' sauthorial intentions-which I cannot justify on thebasis of the evidence so far presented.

VIastos 's developn1ent of ten opposing theses to distinguish SocratesE fron1SocratesM ought to have paid off in a far clearer picture than it hase Of the tenelenctic dialogues, only the Hippias Minor remains free of 'middle' elements inmy account;27 the transitional dialogues flip-flop from early to middle and backagain to meet the requirements of particular theses; and none of the seven dia­logues of group II appears untarnished by 'early' residue. I can only concludethat what supports VIastos 's chronological structure is its scaffolding.28

University of the Witwatersrand(1080 Knapp Street NEGrand Rapids MI 49505)

BIBLIOGRAPHY

Allen, R.E. 1970. Plato's 'Euthyphro , and the Earlier Theory 0/Forms. London: Routledge & KeganPau1.

Benardete, S. 1991. The Rhetoric 0/ Morality and Philosophy: Plato's Gorgias and Phaedrus.Chicago: University of Chicago Press.

Burnyeat, M. 1977. 'Socratic Midwifery, Platonic Inspiration' Bulletin 0/ the Institute 0/ ClassicalStudies 24: 7-15.

Campbell, L. 1867. The Sophistes and Politicus 0/ Plato. Oxford: Clarendon Press.Crombie, I.M. 1962-1963. An Examination 0/Plato's Doctrines. 2 vols. London: Routledge & Kegan

Paul.Davidson, D. 1985. 'P1ato's Philosopher' London Review 0/Books (l August), 15-17.Dodds, E.R. 1959. Plato's Gorgias. Oxford: C1arendon Press.Findlay, J.N. 1974. Plato: The Written and Unwritten Dialogues. New York: Humanities Press.Gundert, H. 1977. Platonstudien. Studien zur Antik Philosophie 7, ed. Klaus Döring and Fe1ix

Preishofen. Amsterdam: B. R. Gruner.How1and, Jacob 1991. 'Re-reading P1ato: The Problem of Platonic Chronology' Phoenix 45: 189­

214.Irwin, T.H. 1977. Plato's Moral Theory; The Early and Middle Dialogues. Oxford: Clarendon Press.Irwin, T.H. 1992. 'The Intellectual Background' Pp. 51-89 in R. Kraut ed. The Cambridge Compan­

ion to Plato. Cambridge: Cambridge University Press.Kahn, C.H. 1981. 'Did Plato Write Socratic Dialogues?' Classical Quarterly 31: 305-320.Kahn, C.H. 1988. 'On the Relative Date of the Gorgias and the Protagoras' Ox/ord Studies in

Ancient Philosophy 6: 69-102.Kahn, C.H. 1992. 'Vlastos's Socrates' Phronesis 37: 233-258.Ledger, G.R. 1989. Re-Counting Plato: A Computer Analysis of Plato's Style. Oxford: Oxford Uni­

versity Press.Martin, G. 1973. Platons Ideenlehre. Berlin and New York: de Gruyter.McPherran, M.L. 1986. 'Socrates and the Duty to Phi1osophize' Southern Journal 0/ Philosophy 24:

541-560.Nails, D. 1992. 'Platonic Chronology Reconsidered' Bryn Mawr Classical Review 3: 314-3~7.

27 And there is doubt about that. Thesleff cautions me, 'It can be argued that Hippias Minor con­tains reminiscences of Gorgias and that its playfu1 irony presupposes a rather sophisticated epistemo­logica1 background (e.g. 375d-e).'

28 I have benefitted from comments of W. Levitan, Holger Thesleff, and anonymous referees.

291

Nails, D. forthcoming. 'Plato's "Middle" Cluster' Phoenix 47.Nehamas, A. 1992. 'Voices of Silence: On Gregory VIastos' Socrates' Arion 2: 157-186.Regenbogen, O. 1950. 'Bemerkungen zur Deutung des platonischen Phaidros' Miscellanea Acad.

Berol.II 1: 198-219.Robinson, R. 1953. Plato's Earlier Dialectic. 2nd edition. Oxford: Clarendon Press.Ross, W.D. 1933. 'The Socratic Problem' Proceedings ofthe Classical Association 30: 7-24.Santas, G.X. 1972. 'The Socratic Fallacy' Journal ofthe History of Philosophy 10: 127-141.Shorey, P. 1903. The Unity ofPlato's Thought. Chicago: University ofChicago Press.Sider, D. 1991. 'Did Socrates Call Himself a Midwife? The Evidence of Clouds' Pp. 333-338 in KJ.

Boudouris ed. Socratic Philosophy. Athens: International Center for Greek Philosophy and Cul­ture.

Stenzei, J.A.H. 1916. 'Literarische Form und philosophische Gehalt des platonischen Dialoges'Jahresbericht der Schlesischen Gesellschaft für vaterländische Kultur. Breslau. Reprinted inKleine Schriften zur Griechischen Philosophie, by Hermann Gentner Verlag, Darmstadt, 1957,32-47.

Stenzei, lA.H. 1917. Studien zur Entwicklung der platonischen Dialektik von Sokrates zu Aristoteles:Arete und Diairesis. Breslau: Trewendt and Granier.

Stenzei, J.A.H. 1940. Plato's Method ofDialectic. Tran. and ed. D.J. Allan. Oxford: Clarendon Press.Teloh, H. 1986. Socratic Education in Plato's Early Dialogues. Notre Dame: University of Notre

Dame Press.Thesleff, H. 1982. Studies in Platonic Chronology. Comnlentationes Hunlanarum Litterarum 70.

Helsinki: Societas Scientiarum Fennica.VIastos, G. 1983a. 'Classical Greek Political Thought: I The Historical Socrates and Athenian

Democracy' Political Theory 11: 495-516.VIastos, G. 1983b. 'The Socratic Elenchus' Oxford Studies in Ancient Philosophy I: 27-58.VIastos, G. 1983c. 'Afterthoughts on the Socratic Elenchus' Oxford Studies in Ancient Philosophy I:

71-74.VIastos, G. 1984. 'Happiness and Virtue in Socrates' Moral Theory' Proceedings of the Cambridge

Philological Society 210 (n.s. 30): 181-213.VIastos, G. 1985. 'Socrates's Disavowal of Knowledge' Philosophical Quarterly 35: 1-31.VIastos, G. 1988a. 'Elenchus and Mathematics: A Turning-Point in Plato's Philosophical Develop-

ment' American Journal ofPhilology 109: 362-396.VIastos, G. 1988b. 'Socrates' Proceedings ofthe British Academy 74: 87-109.VIastos, G. 1990. 'Is the "Socratic Fallacy" Socratic?' Ancient Philos~phy 10: 1-16.VIastos, G. 1991. Socrates: Ironist and Moral Philosopher. Cornell Studies in Classical Philology 50.

Cambridge: Cambridge University Press.Woodruff, P. 1982. Plato, 'Hippias Major'. Indianapolis: Hackett.